| 92 | 0 | 358 |
| 下载次数 | 被引频次 | 阅读次数 |
随着新型能源渗透率的迅速提升和柔性互联设备的广泛接入，中压配电网结构逐渐趋于频繁的动态变化特性，导致传统故障研判方法难以适应。为此，本文提出一种基于多核支持向量数据描述（SVDD）的故障研判方法。首先，通过TOPSIS-Kmedoids算法对基础拓扑特征进行筛选，根据其对故障研判的影响程度赋予权重形成拓扑敏感特征集；接着引入多核学习技术，将拓扑敏感特征与量测采集深度融合，进一步增强模型对节点异常特征的检测能力并训练SVDD模型；最后，充分考虑拓扑变动情况，采用最短路径核策略对新时戳下SVDD模型的超球边界自适应训练调整，分离存在异常的故障节点。所提方法已在华东某电力公司试点部署，并利用实际配电网数据校核验证，研判准确率达到95.8%，具备良好的鲁棒性和可行性。
Abstract：With the rapid increase in the penetration rate of new energy sources and the wide access of flexible interconnection devices,the structure of medium voltage distribution network gradually tends to frequent dynamic changes in characteristics,resulting in the traditional fault investigation and determination methods are difficult to adapt.To this end,this paper proposes a fault investigation method based on multi-core support vector data description(SVDD).Firstly,the basic topological features are screened by TOPSIS-Kmedoids algorithm,and the topology-sensitive feature set is formed by assigning weights according to the degree of its influence on fault investigation.Then,multi-core learning technology is introduced to deeply integrate the topology-sensitive features with the measurement collection to further enhance the model's ability of detecting the abnormal features of the nodes and to train the SVDD model.Finally,the shortest-path kernel strategy is adopted by taking the topological changes into full consideration.The shortest path kernel strategy is used to adaptively train the hypersphere boundaries of the SVDD model under the new time-stamp to separate the faulty nodes with anomalies.The proposed method has been deployed in a pilot project of a power company in East China,and verified with actual distribution network data.The accuracy rate reaches 95.8% which has good robustness and feasibility.
[1]王成山，王瑞，冀浩然，等.多电压等级配电系统智能软开关协同配置[J].中国电机工程学报，2024,44(17):6831-6844.Wang Chengshan,Wang Rui,Ji Haoran, et al. Coordinated Allocation of SOP in Multi-voltage Distribution Network[J]. Proceedings of the CSEE,2024,44(17):6831-6844.
[2]胡鹏飞，朱乃璇，江道灼，等.柔性互联智能配电网关键技术研究进展与展望[J].电力系统自动化，2021,45(8):2-12.Hu Pengfei,Zhu Naixuan,Jiang Daozhuo,et al.Progress and Prospects of Flexible Interconnected Smart Distribution Grid Key Technology Research[J]. Automation of Electric Power Systems,2021,45(8):2-12.
[3]Jiang X,Zhou Y,Ming W,et al. An Overview of Soft Open Points in Electricity Distribution Networks[J].IEEE Trans. on Smart Grid,2022,13(3):1899-1910.
[4]骆晨，冯玉，吴凯，等.基于多源停电数据提示学习的电网轻量化停电感知模型[J].现代电力，2025,40(3):421-431.Luo Chen,Feng Yu,Wu Kai,et al.Light Weight Outage Sensing Model for Grid with Cue Learning From Multiple Outage Data Sources[J].Modern Electric Power,2025,40(3):421-431.
[5]Teng J,Huang W,Luan S. Automatic and Fast Faulted Line-section Location Method for Distribution Systems Based on Fault Indicators[J].IEEE Trans. on Power Systems,2014,29(4):1653-1662.
[6]麦章渠，曾颖，张禄亮，等．基于改进哈里斯鹰优化算法的有源配电网故障定位[J].智慧电力，2022,50(11):104-111.Mai Zhangqu,Zeng Ying,Zhang Luliang,et al. Active Distribution Network Fault Location Based on Improved Harris Hawk Optimization Algorithm[J]. Smart Power,2022,50(11):104-111.
[7]郭壮志，陈波，刘灿萍，等.基于遗传算法的配电网故障定位[J].电网技术，2007,31(11):88-92.Guo Zhuangzhi,Chen Bo,Liu Canping,et al.Fault Location of Distribution Network Based on Genetic Algorithm[J].Power System Technology,2007,31(11):88-92.
[8]颜景斌，夏赛，王飞，等.基于改进遗传算法的有源配电网故障定位分析[J].电力系统及其自动化学报，2019,31(6):107-112.Yan Jingbin,Xia Sai,Wang Fei,et al.Analysis of Fault Location for Active Distribution Network Based on Improved Genetic Algorithm[J]. Proceedings of the CSUEPSA,2019,31(6):107-112.
[9]Wang Y,Wang T,Liu L. A Fault Segment Location Method for Distribution Networks Based on Spiking Neural P Systems and Bayesian Estimation[J]. Protection and Control of Modern Power Systems,2023,8(3):783-794.
[10]王鹤，韦搏，李石强，等.基于贝叶斯网络信息融合的直流配电网故障诊断方法[J].电力系统保护与控制，2024,52(5):61-72.Wang He,Wei Bo,Li Shiqiang,et al. Fault Diagnosis Method for DC Distribution Network Based on Bayesian Network Information Fusion[J]. Power System Protection and Control,2024,52(5):61-72.
[11]潘思宇，刘宝柱.面向配电网故障快速处理的边缘计算单元优化配置方法[J].电力建设，2022,43(3):31-41.Pan Siyu,Liu Baozhu.An Optimal Allocation Method of Edge Computing Units for Fast Fault Processing in Distribution Networks[J].Electric Power Construction,2022,43(3):31-41.
[12]魏然，张磐，高强伟，等.基于网络树状图的低压配电网故障研判仿真分析[J].电力系统保护与控制，2021,49(13):167-173.Wei Ran,Zhang Pan,Gao Qiangwei,et al. Simulation Analysis of Fault Investigation and Determination of Low Voltage Distribution Network Based on Network Tree Diagrams[J].Power System Protection and Control,2021,49(13):167-173.
[13]Luo J,Liu Y,Cui Q,et al. Single-ended Time Domain Fault Location Based on Transient Signal Measurements of Transmission Lines[J]. Protection and Control of Modern Power Systems,2024,9(2):61-74.
[14]Chen R,Li X,Chen Y.Optimal Layout Model of Feeder Automation Equipment Oriented to the Type of Fault Detection and Local Action[J]. Protection and Control of Modern Power Systems,2023,8(1):15-29.
[15]Tang M,Lu H,Li B. Fault Location of Untransposed Double-circuit Transmission Lines Based on an Improved Karrenbauer Matrix and the QPSO Algorithm[J]. Protection and Control of Modern Power Systems,2023,8(3):738-751.
[16]李晓露，陈思羽，柳劲松，等.基于数据增强的配电网运行态势预警方法[J].电网技术，2024,49(3):1207-1216.Li Xiaolu,Chen Siyu,Liu Jinsong,et al.A Data Augmentation Based Approach for Early Warning of Distribution Network Operation Posture[J].Power System Technology,2024,49(3):1207-1216.
[17]Li D,Zhao Y,Zhao Y. A Dynamic-model-based Fault Diagnosis Method for a Wind Turbine Planetary Gearbox Using a Deep Learning Network[J].Protection and Control of Modern Power Systems,2022,7(2):323-337.
[18]孙鹏，陈皓，蒋莉，等.柔性配电网中基于多源信息融合与自适应核学习的故障区段定位[J].电力系统保护与控制，2023,51(7):1-10.Sun Peng,Chen Hao,Jiang Li,et al.Fault Section Location in Flexible Distribution Networks Based on Multi-Source Information Fusion and Adaptive Kernel Learning[J].Power System Protection and Control,2023,51(7):1-10.
[19]赵洪山，王鹏，邵华，等.基于拓扑演变与多核SVDD的主动配电网异常检测方法[J].电工技术学报，2021,36（增刊1）:98-107.Zhao Hongshan,Wang Peng,Shao Hua,et al. Anomaly Detection in Active Distribution Networks Based on Topology Evolution and Multi-kernel SVDD[J]. Transactions of China Electrotechnical Society,2021,36(S1):98-107.
[20]Cheng L,Wang T,Wang Y. A Novel Fault Location Method for Distribution Networks with Distributed Generations Based on the Time Matrix of Traveling-wave[J].Protection and Control of Modern Power Systems,2022,7(4):698-708.
基本信息:
DOI：10.20222/j.cnki.cn61-1124/tm.20250509.015
中图分类号:TM73
引用信息:
[1]杨坤,鲍若愚,徐斌.基于多核SVDD的柔性配电网故障研判方法[J].电力电子技术,2026,60(03):97-108.DOI:10.20222/j.cnki.cn61-1124/tm.20250509.015.
基金信息:
国家自然科学基金（51907114）